转载：
http://www.blogjava.net/hwpok/archive/2008/07/25/217448.html
假如有一张users表,里面有user_id,user_name两个字段,用一条sql语句去掉数据库里的userName重复的记录.
1.delete from users a where a.user_id = (select max(b.user_id) from users b where a.user_name=b.user_name);
这一种自联接,效率低一点.
2.delete from users where user_id not in(select max(user_id) from users group by user_name having count(user_name)>=1);
这一种先分组,效率高一些.
在此作下符号,希望对同行的朋友们有所启示.
| 继往开来 吐故纳新 |
|
||||||
|
||||||